TenFourFox 45.4.0 available (plus: priorities for feature parity and down with Dropbox)

With the peerless level of self-proclaimed courageousness necessary to remove a perfectly good headphone jack for a perfectly stupid reason (because apparently no one at Apple charges their phone and listens to music at the same time), TenFourFox 45.4.0 is released (downloads, hashes, release notes). This will be the final release in the beta cycle and assuming no critical problems will become the public official release sometime Monday Pacific time. Localizations will be frozen today and uploaded to SourceForge sometime this evening or tomorrow, so get them in ASAP.

The major change in this release is additional tweaking to the MediaSource implementation and I'm now more comfortable with its functioning on G4 systems through a combination of some additional later patches I backported and adjusting our own hacks to not only aggressively report the dropped frames but also force rebuffering if needed. The G4 systems now no longer seize and freeze (and, occasionally, fatally assert) on some streams, and the audio never becomes unsynchronized, though there is some stuttering if the system is too overworked trying to keep the video and audio together. That said, I'm going to keep MediaSource off for 45.4 so that there will be as little unnecessary support churn as possible while you test it (if you haven't already done so, turn media.mediasource.enabled to true in about:config; do not touch the other options). In 45.5, assuming there are no fatal problems with it (I don't consider performance a fatal flaw, just an important one), it will be the default, and it will be surfaced as an option in the upcoming TenFourFox-specific preference pane.

However, to make the most of MediaSource we're going to need AltiVec support for VP9 (we only have it for VP3 and VP8). While upper-spec G5 systems can just power through the decoding process (though this might make hi-def video finally reasonable on the last generation machines), even high-spec G4 systems have impaired decoding due to CPU and bus bandwidth limitations and the low-end G4 systems are nearly hopeless at all but the very lowest bitrates. Officially I still have a minimum 1.25GHz recommendation but I'm painfully aware that even those G4s barely squeak by. We're the only ones who ever used libvpx's moldy VMX code for VP8 and kept it alive, and they don't have anything at all for VP9 (just x86 and ARM, though interestingly it looks like MIPS support is in progress). Fortunately, the code was clearly written to make it easier for porters to hand-vectorize and I can do this work incrementally instead of having to convert all the codec pieces simultaneously.

Interestingly, even though our code now faithfully and fully reports every single dropped frame, YouTube doesn't seem to do anything with this information right now (if you right-click and select "Stats for nerds" you'll see our count dutifully increase as frames are skipped). It does downshift for network congestion, so I'm trying to think of a way to fool it and make dropped frames look like a network throughput problem instead. Doing so would technically be a violation of the spec but you can't shame that which has no shame and I have no shame. Our machines get no love from Google anyway so I'm perfectly okay with abusing their software.

I have the conversion to platform codec of our minimp3 decoder written as a first draft, but I haven't yet set that up or tested it, so this version still uses the old codec wrapper and still has the track-shifting problem with Amazon Music. That is probably the highest priority for 45.5 since it is an obvious regression from 38. On the security side, this release also disables RTCPeerConnection to eliminate the WebRTC IP "leak" (since I've basically given up on WebRTC for Power Macs). You can still reenable it from about:config as usual.

The top three priorities for the next couple versions (with links to the open Github issues) are, highest first, fixing Amazon Music, AltiVec VP9 codepaths and the "little endian typed array" portion of IonPower-NVLE to fix site compatibility with little-endian generated asm.js. Much of this work will proceed in parallel and the idea is to have a beta 45.5 for you to test them in a couple weeks. Other high priority items on my list to backport include allowing WebKit/Blink to eat the web supporting certain WebKit-prefixed properties to make us act the same as regular Firefox, support for ChaCha20+Poly1305, WebP images, expanded WebCrypto support, the "NV" portion of IonPower-NVLE and certain other smaller-scope HTML/CSS features. I'll be opening tracking issues for these as they enter my worklist, but I have not yet determined how I will version the browser to reflect these backported new features. For now we'll continue with 45.x.y while we remain on 45ESR and see where we end up.

As we look into the future, though, it's always instructive to compare it with the past. With the anticipation that even Google Code's Archive will be flushed down the Mountain View memory hole (the wiki looks like it's already gone, but you can get most of our old wikidocs from Github), I've archived 4.0b7, 4.0.3, 8.0, 10.0.11, 17.0.11 and Intel 17.0.2 on SourceForge along with their corresponding source changesets. These Google Code-only versions were selected as they were either terminal (quasi-)ESR releases or have historical or technical relevance (4.0b7 was our first beta release of TenFourFox ever "way back" in 2010, 8.0 was the last release that was pure tracejit which some people prefer, and of course Intel 17.0.2 was our one and so far only release on Intel Macs). There is no documentation or release notes; they're just there for your archival entertainment and foolhardiness. Remember that old versions run an excellent chance of corrupting your profile, so start them up with one you can throw away.

Finally, a good reason to dump Dropbox (besides the jerking around they give those of you trying to keep the PowerPC client working) is their rather shameful secret unauthorized abuse of your Mac's accessibility framework by forging an entry in the privacy database. (Such permissions allow it to control other applications on your Mac as if it were you at the user interface. The security implications of that should be immediately obvious, but if they're not, see this discussion.) The fact this is possible at all is a bug Apple absolutely must fix and apparently has in macOS Sierra, but exploiting it in this fashion is absolutely appalling behaviour on Dropbox's part because it won't even let you turn it off. To their credit they're taking their lumps on Hacker News and TechCrunch, but accepting their explanation of "only asking for privileges we use" requires a level of trust that frankly they don't seem worthy of and saying they never store your administrator password is a bit disingenuous when they use administrative access to create a whole folder of setuid binaries -- they don't need your password at that point to control the system. Moreover, what if there were an exploitable security bug in their software?

Mind you, I don't have a problem with apps requesting that access if I understand why and the request isn't obfuscated. As a comparison, GOG.com has a number of classic DOS games I love that were ported for DOSBox and work well on my MacBook Air. These require that same accessibility access for proper control of input methods. Though I hope they come up with a different workaround eventually, the GOG installer does explain why and does use the proper APIs for requesting that privilege, and you can either refuse on the spot or disable it later if you decide you're not comfortable with it. That's how it's supposed to work, but that's not what Dropbox did, and they intentionally hid it and the other inappropriate system-level things they were sneaking through. Whether out of a genuine concern for user experience or just trying to get around what they felt was an unnecessary security precaution, it's not acceptable and it's potentially exploitable, and they need to answer for that.

Watch for 45.4 going final in a couple days, and hopefully a 45.5 beta in a couple weeks.

System7SoonToBeYesterday

One of my periodic "drop by now and then" sites is System7Today, extolling the virtues of System 7 to those people still rocking 60x-series Power Macs (that's Mac OS 7.0 through 7.6.1 for those of you who only know Mac OS as starting with a lower-case m, and also, get off my lawn), and I was disheartened to see that the System7Today forums are going read-only. I regretfully understand his reasoning though one wonders what will happen to the rest of the site. It hasn't changed in years, but it hasn't had to, and I love the frozen-in-time mockery of the Apple front page from at least a decade or so prior.

The only Macs I have still running System 7 are all running 7.1 (not counting the NetBSD 68K systems, which just use System 7 as a bootloader): one is my IIci with a 50MHz DayStar '030 and a MacIvory Lisp card, another is my recapped SE/30 looking for a job to do, and the last is my super-cute Mystic Color Classic. I like 7.1 a lot more than 7.5 or 7.6, and you can transplant lots of the 7.5 CDEVs and such back to 7.1 for a slimmer but still feature-filled experience. That said, I have to confess that I jumped to OS 8, and then OS 9, whenever I got the chance on my Power Macs. Part of this was that I upgraded those systems aggressively -- all of my Old World Power Macs in regular use have G3 or G4 upgrade cards, including my beloved PowerBook 1400, so they all run 9.1 or 9.2.2 -- and part of it is to run Classilla, but the biggest reason was just that OS 8 looked nicer and felt better, and 8.1 and 8.6 were still pretty speedy. I still run OS 8 on my PowerBook 2300c and Quadra 800.

But still, nostalgia dies hard. (No doubt being a TenFourFox user, you'd empathize.) While I've got lots of classic System 7-era software backed up for posterity on the Floodgap gopher server, it just doesn't have System7Today's throwback vibe and playful attitude. I'll miss it dearly pending its inevitable slow-motion decommissioning, most of all because it remains a great example of doing the most you can with the little you've got. Typing this blog post on a 2002-vintage 1GHz iMac G4, us PowerPC holdouts are still living in that spirit today.

Watch for 45.4 later this week(end).

Talos moves to CrowdSupply

It looks like we've cleared the hump: Raptor Engineering tells me that Talos, the upcoming POWER8 workstation and probably the most powerful non-x86 workstation-class system you can buy, has enough interest to move to a formal CrowdSupply launch. Sign up and stay tuned. I've been still saving my pennies and I'm definitely buying.

TenFourFox 45.3.0b2 available

TenFourFox 45.3 beta 2 is now available (downloads, hashes, release notes). The issue with video frame skipping gone awry seems to be due to Mozilla's MediaSource implementation, which is (charitably) somewhat incomplete on 45 and not well-tested on systems of our, uh, vintage. Because audio is easier to decode than video the audio decoder thread will already have a speed advantage, but in addition the video decoder thread appears to only adjust itself for the time spent in decoding, not the time spent pushing the frames to the compositor or blitting them to the screen (which are also fairly costly in our software-rendered world). A frame that's simple to decode but requires a lot of painting will cause the video thread to get further and further behind queueing frames that the compositor can't keep up with. Even if you keep skipping keyframes, you'll end up more than one keyframe behind and you'll never catch up, and when this occurs the threads irrevocably go out of sync. The point at which this happens for many videos is almost deterministic and correlates well with time spent in compositing and drawing; you can invariably trigger it by putting the video into full screen mode which dramatically increases the effort required to paint frames. In a few cases the system would just give up and refuse to play further (hitting an assertion in debug mode).

After a few days of screaming and angry hacking (backporting some future bug fixes, new code for dumping frames that can never be rendered, throwing out the entire queue if we get so far behind we'll never reasonably catch up) I finally got the G5 to be able to hold its own in Reduced Performance but it's just too far beyond the G4 systems to do so. That's a shame since when it does work it appears to work pretty well. Nevertheless, for release I've decided to turn MediaSource off entirely and it now renders mostly using the same pipeline as 38 (fortunately, the other improvements to pixel pushing do make this quicker than 38 was; my iBook G4/1.33 now maintains a decent frame rate for most standard definition videos in Automatic Performance mode). If the feature becomes necessary in the future I'm afraid it will only be a reasonable option for G5 machines in its current state, but at least right now disabling MediaSource is functional and sufficient for the time being to get video support back to where it was.

The issue with the Amazon Music player is not related to or fixed by this change, however. I'm thinking that I'll have to redo our minimp3 support as a platform MP3 decoder so that we use more of Mozilla's internal media machinery; that refactoring is not likely to occur by release, so we'll ship with the old decoder that works most other places and deal with it in a future update. I don't consider this a showstopper unless it's affecting other sites.

Other changes in 45.3 beta 2 are some additional performance improvements. Most of these are trivial (including some minor changes to branching and proper use of single-precision math in the JIT), but two are of note, one backport that reduces the amount of allocation with scrolling and improves overall scrolling performance as a result, and another one I backported that reduces the amount of scanning required for garbage collection (by making garbage collection only run in the "dirty" zones after cycle collection instead of effectively globally). You'll notice that RAM use is a bit higher at any given moment, but it does go back down, and the browser has many fewer random pauses now because much less work needs to be done. Additionally in this beta unsigned extensions should work again and the update source is now set to the XML for 45. One final issue I want to repair before launch is largely server-side, hooking up the TenFourFox help button redirector to use Mozilla's knowledge base articles where appropriate instead of simply redirecting to our admittedly very sparse set of help documents on Tenderapp. This work item won't need a browser update; it will "just work" when I finish the code on Floodgap's server.

At this point barring some other major crisis I consider the browser shippable as 45.4 on time on 13 September, with Amazon Music being the highest priority issue to fix, and then we can start talking about what features we will add while we remain on feature parity. There will also be a subsequent point release of TenFourFoxBox to update foxboxes for new features in 45; stay tuned for that. Meanwhile, we could sure use your help on a couple other things:

• Right now, our localized languages for the 45 launch are German, Spanish, French and Japanese, with an Italian translation in progress. We can launch with that, but we're losing a few languages there we had with 38. You can help.

• Even minor releases of the browser get a small bump in support tickets on Tenderapp; a major release (such as 31 to 38, and now 38 to 45) generally brings lots of things out of the woodwork that we've never seen during the beta cycle. Most of these reports are spurious, but sometimes serious bugs have been uncovered even though they looked obviously bogus at first glance, and all of them need to be triaged regardless. If you're a reasonably adept user and you've got a decent understanding of how the browser works, help us help your fellow users by being a volunteer on Tenderapp. There's no minimum time commitment required; any help you can offer is appreciated, even if just temporary. Right now it's just Chris Trusch and me trying to field these reports, which is unsatisfactory to cover all of them (Theo used to but we haven't seen him in awhile). If you'd like to help, post in the comments and we'll make you a "blue name" on Tenderapp too.

That's it for now. See what you think.

TenFourFox 45 beta 2: not yet

So far the TenFourFox 45 beta is doing very well. There have been no major performance regressions overall (Amazon Music notwithstanding, which I'll get to presently) and actually overall opinion is that 45 seems much more responsive than 38. On that score, you'll like beta 2 even more when I get it out: I raided some more performance-related Firefox 46 and 47 patches and backported them too, further improving scrolling performance by reducing unnecessary memory allocation and also substantially reducing garbage collection overhead. There are also some minor tweaks to the JavaScript JIT, including a gimme I should have done long ago where we use single-precision FPU instructions instead of double-precision plus a rounding step. This doesn't make a big difference for most scripts, but a few, particularly those with floating point arrays, will benefit from the reduction in code size and slight improvement in speed.

Unfortunately, while this also makes Amazon Music's delay on starting and switching tracks better, it's still noticeably regressed compared to 38. In addition, I'm hitting an assertion with YouTube on certain, though not all, videos over a certain minimum length; in the release builds it just seizes up at that point and refuses to play the video further. Some fiddling in the debugger indicates it might be related to what Chris Trusch was reporting about frameskipping not working right in 45. If I'm really lucky all this is related and I can fix them all by fixing that root problem. None of these are showstoppers but that YouTube assertion is currently my highest priority to fix for the next beta; I have not solved it yet though I may be able to wallpaper it. I'm aiming for beta 2 this coming week.

On the localization front we have French, German and Japanese translations, the latter from the usual group that operates separately from Chris T's work in issue 42. I'd like to get a couple more in the can before our planned release on or about September 13. If you can help, please sign up.

And now for something completely different: what you missed by not going to Vintage Computer Festival West XI

With the exception of this man,

you all missed out by not coming to the reincarnated Vintage Computer Festival West. You had a chance to meet me and my beloved Tomy Tutor (which you can emulate on a Power Mac or Intel Mac). You had a chance to see other inferior exhibits. You didn't. So let me now make you feel bad about what you missed out on.

You could have seen the other great exhibits at the Computer History Museum,

the complete Tomy Tutor family, hosted by my lovely wife and me

that corrupted young minds from a very early age

(for which I won a special award for Most Complete and 2nd place overall in the Microcomputer category),

seen a really big MOS 6502 recreation,

played Maze War,

solved differential equations the old fashioned way,

realized that in Soviet Russia computer uses you,

given yourself terrible eyestrain,

used a very credible MacPaint clone on a Tandy Color Computer,

messed with Steve Chamberlain's mind,

played with an unusual Un*x workstation powered by the similarly unusual (and architecturally troublesome) Intel i860,

marveled at the very first Amiga 1000 (serial #1),

attempted to steal an Apple I,

snuck onto Infinite Loop on the way back south,

and got hustled off by Apple security for plastering "Ready for PowerPC upgrade" stickers all over the MacBooks in the company store.

But you didn't. And now it's too late.^*

^*Well, you could always come next year.

45.3 beta 1 available

For better or worse, I bring you fire TenFourFox 45.3 beta 1 (downloads, hashes, release notes).

Like pretty much every major ESR jump from before, some sites will be faster and some sites will be slower. The biggest new change is that we are now building the full Intl API with ICU support (issue 304). We were able to get away without it in 31 and 38 due to various hacked-in stubs but now too much in the browser (let alone add-ons) requires it. To avoid issue 266 we build a shared ICU and shared data file which will degrade startup time while they get searched and loaded along with all the others. There are also new UI features in the browser such as search suggestions (which you can disable) that add overhead, and certain layout and rendering operations are more heavyweight than they were in earlier releases. I've tried to load in some later performance updates to counter this last issue, but they don't erase the delta completely. On the plus side, however, as mentioned previously JavaScript JIT performance jumped 20% between 38 and 45, and low-level graphics performance and software pixel pushing is faster which manifests in better animation performance and smoother scrolling.

Remaining issues include Amazon Music failing to advance tracks in playlists (the JIT has been exonerated, so I'm still spinning my wheels here), disabling the last remnants of signed extensions (which I philosophically disagree with) and changing the update notification source. Otherwise it seems to function very well once it's loaded everything and got most of the browser chrome in the JIT. The browser is tested on my usual lab machines, the 16GB monster Quad G5, a 1.5GB 1.33GHz iBook G4, a 1.25GB 1GHz iMac G4 15" and a 2GB 450MHz Sawtooth G4, all running Tiger. TenFourFoxBox appears to work with no changes needed but you may want to keep 38.10 around for right now if you are using foxboxes heavily, just in case (please report issues as you find them).

On the build side, pretty much the process is the same as 38 except that gcc 4.8 is now required. The MacPorts build of gcc is what I use, and I'd like to hear if other builds from our other unofficial package maintainers also work. The hacked strip7 unfortunately is still needed for stripping release binaries to handle the variant N_SECTs in the object files. One irregularity is that gcc 4.8 can't distinguish Objective-C messages from C++11 lambdas, or rather, that Objective-C++ has priority over lambdas, so that you can't reliably use both language features in the same file. This occurs in exactly one place in the entire code base, in code we don't support anyway, so it's #ifdefed off. If your local changes run afoul of this, try adding #define GCC_SUCKS_MOOSE_WANG (yes, really, I was a bit peeved that day) and see if it builds with that, or put the C++11 lambdas into another file and have your Objective-C++ call it instead. The build instructions are updated and available now, and with luck our esteemed colleague in the Land of the Rising Sun will be able to build Tenfourbird on the 45 base as well. Also, as you will see, the G3 version built just fine and will still be supported in 45 and thereafter.

Localizers should be aware of two new files, TenFourFox.dtd and TenFourFox.properties, which will be the locale basis for features I am considering or planning to release. I'd rather have translations for features I decide against implementing than not having strings for features I do implement, so almost nothing in those two files (which are roughly the same strings in different places) is being used right this second; that said, however, the idea is to roll out features using these strings during 45's lifetime as we move from source parity to feature parity and this will enable me to do so more quickly. If you would like to contribute to the localization effort (we really appreciate it), Chris T wrangles the cats in issue 42. The aim is to have substantially the same languages available for 38 available for 45 at or near the time of release in about six weeks. Your help is invaluable in this effort.

Finally, let's give 38 a rousing round of applause for being a solid, long lived branch release as it rides off into the ESR sunset. Its end also means the end of one of the oddest chapters in Firefox history, Debian Iceweasel, which has now reverted to being "just" Firefox as of 45ESR. Although both we and they rebranded ourselves because of various changes to Firefox we were making, we'll still be keeping our name because we really do change the feature set and make various platform-specific changes Mozilla would probably find unacceptable otherwise. Plus it gives you guys something to still complain about, along with the icon. :)

Don't file Github tickets on beta issues; post them in the comments. Please be kind. This was literally months of work. Thanks and let's start banging the bugs out.